Re: [boost] Re: Re: Re: Re: Profiling Library suggestion

From: Rob Stewart (stewart_at_[hidden])
Date: 2005-02-14 16:06:10

From: "Preston A. Elder" <prez_at_[hidden]>
>
> I don't know of any windows commandline utilities to track this, but If
> you, say, had an application on linux/unix and ran 'time' on it, like so:
> 
> time tar xvfz ../mtxdrivers-rh9.0-v1.1.0-pro-beta.tar.gz
> 
> You would get the following output:
> 
> real    0m0.414s
> user    0m0.166s
> sys     0m0.104s
> 
> The 'real' is how much the difference between the wall clock times at the
> start and end of the task.
> 
> The user and sys are how much time in 'CPU seconds' (aka. how many
> CPU clock cycles) the application was given to perform its task ('user'
> being how much time the user-space portion of the code used, and 'sys'
> being how much time the kernel took to execute its part of it (eg. disk
> I/O, etc).

Note that the times are all reported in seconds.  Whatever
internal type is used to track the values, time converts it to
seconds.

The profiler must support the same thing, and I think it does.
Whatever the timing policy uses as the timing type, the reporting
policy must interpret correctly.  It would be nice if there was a
means to ensure they are assembled correctly so that a report of
elapsed time in seconds isn't mislabeled as clock ticks.

-- 
Rob Stewart                           stewart_at_[hidden]
Software Engineer                     http://www.sig.com
Susquehanna International Group, LLP  using std::disclaimer;